Download - Gear Technology magazine
Download - Gear Technology magazine
Download - Gear Technology magazine
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
creases about 45% for three of the four<br />
cases and for the A6 quality, the stress<br />
increase is still about 20%.<br />
Runout deviations. In the simulations<br />
performed for this paper,<br />
AGMA radial runout was converted<br />
to tangential spacing errors using the<br />
tangent of the pressure angle. The<br />
runout was considered to be sinusoidal<br />
and a sinusoidal train of spacing errors<br />
was simulated. Values are not presented,<br />
but values were roughly equivalent to<br />
those of individual spacing errors of the<br />
“worst- case spacing” of the sinusoidal<br />
errors.<br />
In converting radial runout to<br />
tangential spacing error, the following<br />
equation was used for each gear:<br />
SE = f idT<br />
2 tan ( t<br />
) 360<br />
N T <br />
where,<br />
SE<br />
peak effective spacing error<br />
FidT radial composite tolerance<br />
t<br />
NT<br />
transverse pressure angle<br />
number of teeth<br />
(3)<br />
Summary<br />
An analysis procedure has<br />
been presented that ac counts for<br />
manufacturing accuracies in evaluating<br />
contact stresses, root stresses and load<br />
distribution factors. The same procedure<br />
may be used for other design metrics<br />
such as film thickness, flash temper ature<br />
and transmission error. The increases in<br />
stresses due to profile and lead deviations<br />
are cer tainly significant, being as high<br />
as 26% for the root stresses of the<br />
example quality A8 helical pinion. The<br />
load distribution factor increased 34%<br />
for the same pinion. Spacing variations<br />
provided even higher stress variation,<br />
with root stresses increas ing by as much<br />
as 45% for the A8 spacing error. Although<br />
not shown, contact stresses and<br />
the load distribution factor also increase<br />
significantly when spacing errors are<br />
applied. Also presented in this paper<br />
is a procedure for selecting appropriate<br />
modi fications that compensate for<br />
misalignment and avoid severe corner<br />
contact and tip interference. General<br />
<br />
<br />
Spacing and Runout Deviations<br />
Both of these deviations fall in the<br />
domain of the AGMA quality system,<br />
and so are evaluated using the load<br />
distribution analysis. A scheme for<br />
evaluating these deviations on both<br />
static and dynamic stresses has been<br />
performed (Ref. 15), and this method<br />
could in fact be superimposed upon<br />
the methodolo gy described in this<br />
paper. Here, we shall only per form the<br />
static analysis. It is expected that one<br />
can simply superimpose the spacing<br />
and runout devi ation effects with the<br />
lead and profile deviations, without<br />
introducing much error. So for now, we<br />
will only look at the spacing and runout<br />
deviation effects as separate cases.<br />
Spacing Deviations. Here, we shall<br />
evaluate only the worst case tooth-totooth<br />
spacing deviation and will present<br />
results only for root stresses. There<br />
seem to be two possibilities for creating<br />
a worst-case scenario—the first being<br />
when only one tooth is mispositioned,<br />
and the second when many subse quent<br />
teeth are mispositioned by the deviation<br />
toler ance. Figure 17 shows the first case<br />
and Figure 18 shows the second case.<br />
In each case the position of the first<br />
tooth with an error may be positive<br />
(comes into contact early) or negative<br />
(comes into contact late). The negative<br />
condition is shown in the two figures.<br />
These effects are simulated by<br />
essentially shifting the profiles either<br />
forward or backward, depending upon<br />
the sign of the deviation. For positive<br />
devi ations, the tooth with the error<br />
comes into contact early and carries a<br />
disproportionate share of the load. The<br />
values of errors used are the sum of the<br />
square of the spacing deviations of the<br />
pinion and the gear, respectively.<br />
Figure 19 shows the plots of root<br />
stresses for five locations across the<br />
face width (similar to Figure 3) for a<br />
modified gear tooth without spacing<br />
errors. Figure 20 shows the stresses<br />
after the addition of the spacing error,<br />
and one notes that the stresses at the<br />
edge of the tooth increase significantly.<br />
Table 5 summarizes the stress values<br />
for positive and neg ative errors for each<br />
case for both A6 and A8 accu racies.<br />
For quality A8 gears, the root stress inconclusions<br />
are:<br />
• Spacing E V deviations: E N They T have S a<br />
large effect on contact and root<br />
stresses, mainly due to transverse<br />
load sharing (up to 50%).<br />
• Runout has a relatively small<br />
effect, but lower quality gears<br />
still may cause an increase in<br />
stresses.<br />
• Microgeometry changes (profile,<br />
lead and misalignment deviations)<br />
significantly affect<br />
contact stresses (5–10%) and<br />
root stresses (10–25%).<br />
• It is important to start a<br />
design with reasonable<br />
profile and lead modifications,<br />
since this reduces the variability<br />
in stresses due to inaccuracies.<br />
• The best modifications do not<br />
totally eliminate corner contact<br />
or tip interference, but do reduce<br />
their effect. If one totally<br />
eliminates tip interference,<br />
stresses will be higher than<br />
those for the“best” modifications<br />
shown here.<br />
• Although the increases in<br />
stress values due to manufacturing<br />
variability appear to be<br />
quite large relative to current<br />
rating practice, one could justify<br />
these effects as being part of<br />
the uncertainty that is part<br />
of the design factor of safety<br />
(difference between design<br />
allowable stresses and material<br />
property stresses) in the current<br />
rating practice.<br />
Even though many manufacturing<br />
variables have been considered in this<br />
study, there are still numer ous factors that<br />
are affected by manufacturing accuracy<br />
that will influence stress values. Several,<br />
such as tooth thickness, center distance<br />
variation, outside diameter variation and<br />
surface finish varia tion, have not been<br />
included and certainly have possibilities<br />
for future studies. Fortunately, many<br />
designers run min/max calculations for<br />
these parameters so they do in a way get<br />
considered in their evaluations.<br />
Acknowledgements<br />
The author would like to thank the<br />
many sponsors of the <strong>Gear</strong> and Power<br />
continued<br />
www.geartechnology.com July 2009 GEARTECHNOLOGY 00 59